Comparing Kill Steps in Raw Pet Food: HPP, Bacteriophages, Lactic Acid, Probiotics & more

Buckle in, this is going to be a looong one. Food safety & pathogen control is one of the most controversial topics in the raw feeding world. Regardless of where you stand with whether or not some bacteria can or should be present in raw foods, the regulatory stance on this is clear. The FDA policy is zero-tolerance (we’ve written more on why here ) and that on its own makes it incredibly important for the industry to look at effective methods of pathogen control.

The task at hand is fundamentally difficult from a scientific perspective. You have raw ingredients that come in with some amount of pathogen load which needs to be eliminated BUT in a way that preserves the raw, minimally processed nature and nutrition of the food. In other words, you need to find something that’s “tough on pathogens but gentle on the food”.

We’ve spent A LOT of time and resources evaluating various different technologies and wanted to share some of our learnings thus far. There’s a lot of techniques out there ranging from those that sound like they should work in theory but saw challenges in live applications to ones that showed surprising efficacy.

A few caveats before we get started. This is based on learnings from our experiences, conversations with food safety experts, microbiologists, and internal testing at Viva. Although this is based on our best current knowledge, it is not a scientific review and represents our current perspectives. There may be nuances we haven’t uncovered yet and we’re constantly looking to learn and evolve our perspectives. What didn’t work for us may work for others and vice versa. Every manufacturer also has a different situation and there is absolutely no one size fits all approach. What’s important is that everyone conducts the appropriate validation study on their chosen technique and continues to verify that through a robust finished product testing program. 

Whether or not something “works” isn’t a black or white designation. The devil is in the details and it boils down to a couple key points.

1. How much microbial load is in your raw materials. If your raw materials contain a larger number of pathogen cells, then you will need a correspondingly more lethal technique. The proper approach is to test your raw materials to find a baseline pathogen load, apply a healthy buffer and make sure your kill step reaches that target level of efficacy. We have an ongoing monitoring program to test our raw materials for this exact reason. As an example, if your testing shows that you never have more than 10 CFUs/g of pathogens in your ingredients, you may only need a technology that is capable of eliminating up to 1,000 cfu/g (safety buffer of 100x). However, if you regularly see 100-1,000 CFU/g in your products, then you need something stronger that can eliminate up to 10,000-100,000 CFU/g (safety factor of 100x) and anything less wouldn’t be considered effective on a consistent basis.

2. How you apply the technology. Almost any technology can fail or succeed depending on how you do it. For example, even with cooking, if a food isn’t heated to the correct temperature and for the right amount of time, then it won’t work. But as we’ll get into later on, applying technologies in an effective way can be challenging in practice. Due to factors like the concentration, amount of product contact, or time required and other impacts it can have on product look & feel, some technologies are just not practical.

3. Results of Your Validation Study. A validation study is where you inoculate your product samples with higher levels of pathogens, apply your treatment, and test to ensure the pathogens are no longer detectable. We’ve performed these studies with Salmonella spp, Listeria monocytogenes, and STEC E.coliusing our pathogen eliminating probiotics as the treatment.

This is the regulatory gold standard to be able to say something “works” and it’s important to do it in your products as differences in your product matrix, application technique, or processing steps can all affect the effectiveness of a technology. It’s important that the study tests for presence/absence and not just log reduction since the goal is non-detectable and not just 100 cells fewer.

4. Ongoing Finished Product Testing. This is your reality check and your finished product testing is going to tell you if your process is working on an ongoing basis. If you’re getting presumptive positives regularly, something is off - whether it’s the environment, inconsistencies in applying techniques, raw materials etc. It’s important that you’re also taking enough samples from each batch to catch what you’re looking for if it’s there. Taking 1 vs. 100 samples per batch is a big difference even though you're still "testing" in both cases.

Efficacy: Our testing didn’t find any meaningful log reduction when using a commercially available bacteriophage cocktail. We inoculated samples with 4 logs of Salmonella spp.and then added 1x and 1.5x the recommended bacteriophage concentration and found no statistically significant difference in the ending concentration of Salmonella spp.between the control (blue), 1x (orange), and 1.5x (green). Sampling of each product occurred at mixing (before freezing). Then the product was vacuum packaged and placed into a refrigerator (40°F) for 1 hour before moving to a freezer (-4°F) to be frozen overnight. The next morning, the product was placed back into the refrigerator to thaw and sampled at 24 hours and 48 hours. 

From our research, bacteriophages are highly host specific which means one strain of phage that infects Salmonella species A may not have the right receptors to infect Salmonella species B.There are hundreds of Salmonella species alone and it is difficult to create a phage cocktail that is effective against all strains and continuously develop phages so they’re effective as host cell receptors mutate. Phages are also highly dependent on coming into contact with host cells and are not very mobile by nature which poses a challenge if the bacterial concentration is sparse or if mixing is not even. In general, the ground, 3-dimensional nature of raw pet food poses challenges for any spray on treatment. As an example, bacteriophages perform much better as a spray on to the surface of chicken breast.

Finally, as a technicality, bacteriophages are not “generally recognized as safe” by the FDA for pet food even though it is approved in human foods.

What is it? Product is placed in a water-filled chamber that gets pressurized and the high pressure eliminates pathogens by disrupting cell membranes and protein structures. Companies frequently send products to a 3rd party to HPP since machines cost millions of dollars.

Sensory Differences: When we tested this on our products, we noticed a significant difference in the color, smell, and texture which is a well documented side effect of this process. The product looked more “cooked” since in both cases, the proteins are being denatured either due to heat or pressure. 

Our product after HPP at various pressures. As pressure increases, so does discoloration and changes in texture.

Efficacy: We have not conducted our own validation of HPP yet but literature shows that it can deliver up to a 4-5 log reduction dependent on target pathogen, product matrix, time, temperature, and pressure used in the process. In discussing with experts, HPP is good at delivering a large log reduction (quantitative measurement) but may not be as robust at eliminating the last few pathogen cells and there are instances where injured cells can recover post HPP and remain viable. At the end of the day, presence/absence (qualitative measurement) is what matters rather than a specific log reduction. In use, you could still receive “presumptive positive” pathogen results if the technology isn’t implemented correctly.

Does it Eliminate “Good” Bacteria? Some claim that HPP doesn’t eliminate good bacteria in the food but in all likelihood, it does reduce beneficial bacteria populations. For example, Lactobacillus is a bacterial family that contains both spoilage and probiotic species and HPP is commonly used to extend product shelf-life by reducing spoilage microorganisms. It stands to reason that the lethality against spoilage bacteria would also apply to probiotic organisms given that they’re in the same genus. Overall, we have not seen any indication that probiotic bacteria are more resistant to high pressure in their cellular structure compared to pathogenic organisms.

However, an important question is how much “good” bacteria is present in meat to begin. If the amount of probiotic bacteria had a negligible health effect to begin with, maybe it doesn’t matter that it’s reduced through HPP. 

Tempering: HPP is ideally performed on product at 38-40F because ice crystals in the product will shield bacterial cells. However, most manufacturers are processing frozen ingredients at 28-32F which results in finished product coming off production lines at colder temperatures than ideal for HPP. Especially at commercial scale, it is notoriously difficult to bring the temperature of finished product up from just 30F to 40F consistently. To make matters worse, these products are palletized when they come off the line which makes tempering even more difficult— if you’ve ever tried to defrost a large amount of meat or even a Thanksgiving turkey, you can imagine how hard this would be. 

What is it? Using organic acids as an antimicrobial rinse or spray on ingredients. This is a common practice within the meat industry and most USDA inspected meat have some sort of acid wash applied.

Efficacy: Our testing didn’t find any meaningful log reduction in Salmonella spp. using up to 4% lactic acid inclusion. Higher concentrations of lactic acid could be more effective but would result in significant issues with how the product looked (grey and oxidized). We inoculated samples with 4 logs of Salmonella spp.and tested samples with 2% and 4% inclusion of lactic acid. Sampling of each product occurred at mixing (before freezing). Then the product was vacuum packaged and placed into a refrigerator (40°F) for 1 hour before moving to a freezer (-20°C) to be frozen overnight. The next morning, the product was placed back into the refrigerator to thaw and sampled at 24 hours and 48 hours. 

Acid rinses are frequently used in meat processing plants so why didn’t it work here? The key difference is that acid washes are typically applied to whole carcasses since this approach relies on surface contact to be effective. As mentioned previously, it’s difficult to make sure it comes into contact with all surfaces inside a ground matrix and that’s likely why there was limited efficacy in our testing. Finally, acid washes in USDA meat processors are used for pathogen reduction, not elimination.  

Efficacy: To be honest, we were skeptical of this technology when we first started learning about it. Unlike many of the other approaches listed here, probiotics seemed the least likely to be effective since you normally think of them in the context of gut health and not pathogen control. Still, we wanted to base our decision off of data and invested in running the validation studies on our products inoculated with Salmonella, Listeria monocytogenes,and STEC E.coli.

This chart shows that the specific probiotics we use were able to eliminate 3 logs of Salmonella and importantly, we continued sampling until day 35 and detected no pathogen cell recovery. Another key benefit of probiotics is that the lethality occurs when the product is in final packaging which prevents environmental recontamination. 

Is a 3 log reduction sufficient? This goes back to our earlier point about raw material testing. We’re extremely selective about our sourcing and test each delivery from our supplier for indicator organisms. To date, all of our raw materials have had <10 CFU/g (1 log) of Salmonella so a 3 log reduction is sufficient for us. However, we’re currently still researching and experimenting with methods to achieve an even greater log reduction.

Why Probiotics Work for Us. Probiotics work for us since we are able to source “cleaner” ingredients that come in with a lower microbial load. We regularly test our raw materials and have made several supplier changes to source higher quality, cleaner ingredients. A key benefit of probiotics is that the lethality occurs when the product is in final packaging. The probiotics do their job while they’re in the sealed package so this prevents environmental recontamination.

Finally, you may be wondering if probiotics are dependent on surface contact like other techniques such as acid rinses. The answer is that they are, but to a much lesser degree. Probiotics are live organisms so they are mobile and in addition, one of their mechanisms of action is producing antimicrobial peptides & acids that further penetrate the product. We also monitor how evenly our probiotics are distributed throughout our batches by testing each lot for the probiotics concentration to make sure we’re within range. 

What’s the Catch? We’ll be the first to say, this approach isn’t for everyone. If your raw materials come in with a higher pathogen load, you’re not able to commit to a rigorous testing program, your facility cleaning & sanitation isn’t buttoned up, or you’re not willing to pay a premium for “cleaner” ingredients, this technology likely won’t work. We’ve spent a lot of time & resources researching, testing, and fine-tuning our application and if there’s enough interest, we’ll consider going into more detail on this in the future!

How is this different from fermentation? Technically, the probiotics in our products undergo fermentation since that just refers to the metabolic process by which these good bacteria break down and use energy. However, our process is very different from the fermentation used to produce familiar foods like yogurt, kimchi, kombucha, or kefir.

The probiotics we’re adding have been carefully selected to not produce any detectable changes in the smell, taste, or texture of the food. For example, our probiotics do not contain any strains traditionally used in fermentation that would produce noticeable amounts of gas, acid, or smells. They are specifically selected for their pathogen eliminating capabilities and are different from the strains used in making fermented foods. Our use of probiotics also does not involve adding any dairy or sugar in our recipes that “feed” the probiotics. Batches made with probiotics look, smell, and taste the same as without.

What is it? Exposing food to ionizing radiation which damages the genetic material of pathogens. This is reportedly a pretty effective technique since the electromagnetic waves can easily penetrate into all areas of the product. We haven’t tested this ourselves due to other concerns.

If whole foods have been irradiated, FDA requires that the label bear the radura symbol and the phrase "treated with radiation" or "treated by irradiation. Interestingly, if irradiated ingredients are added to foods that have not been irradiated, no special labeling is required on retail packages. There could be manufacturers that take advantage of this technicality. 

What is it? This was one of our more adventurous ideas at the start. We started off with a test by putting our raw ingredients through a commercial steam machine so just the outside would be heated with the assumption that most pathogens would be on product surfaces. There were several reasons why we didn't pursue this approach further but we'll save that for another time.

Then we found a company that did use steam successfully on a commercial scale to produce a raw product. Their machine heated AND cooled meat particles in less than a blink of an eye which destroyed the pathogens but kept the product raw. Again, there were a few reasons we couldn’t adopt the approach but it’s an absolutely fascinating technology and something we’ll keep an eye on. 

Left: Partially steamed raw ingredient. Right: "Flash" steamed raw turkey "chips"

What is it? These are strong, highly reactive oxidizing compounds commonly used for disinfection. They’re unstable compounds that quickly break down into oxygen (and water).

Efficacy: We haven’t tested this approach ourselves but these compounds are strong, proven disinfectants for surfaces. However, their application in food is more challenging because they require surface contact. As you can imagine, it’s very difficult to thoroughly mix a gas into a batch of ground meat and to do so before the compound decomposes into its harmless constituents. This could work if you just wanted to treat the surface of a product and we’ve most commonly seen this technology used for environmental sanitation.